| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
, ß, and P in Peripheral Blood Mononuclear Leukocytes in Healthy Controls and in Patients with Hyper- and Hypocortisolism
Department of Internal Medicine, Section of Endocrinology (A.H., J.W.K., F.H.d.J., S.W.J.L., R.A.F.), and Department of Reproduction and Development (A.O.B.), Erasmus Medical Center, 3000 CA Rotterdam, The Netherlands
Address all correspondence and requests for reprints to: Dr. Jan W. Koper, Department of Internal Medicine, Section of Endocrinology, Room Ee 585, Erasmus Medical Center, P.O.B. 2040, 3000 CA Rotterdam, The Netherlands. E-mail: f.koper{at}erasmusmc.nl.
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsDiscussionReferences |
|---|
-isoform of the glucocorticoid receptor (GR). GR- levels and activity are modulated by alternative splicing of the common pre-mRNA into mRNAs for the GR-ß and GR-P isoforms. Objective: The objective of this study was to investigate whether chronic hypercortisolism, chronic hypocortisolism, or acute, relative hypocortisolism influences the expression levels of the GR splice variants in mononuclear leukocytes.
Design: This was a case-control study.
Setting: The study was performed at a university hospital.
Participants: Eighteen patients with Cushing’s syndrome, five patients with hypocortisolemia, seven patients undergoing metyrapone testing, and 14 controls were studied.
Main Outcome Measures: The main outcome measures were mRNA levels, GR affinity, and number per cell.
Results: All three GR mRNA isoforms were detected in participants from all groups at relative levels of /P/ß = 1:0.25:0.001. There was a significant correlation between the expression levels of the three splice variants and between the mRNA levels and the number of receptors per cell. The GR in Cushing patients had an increased Kd (P < 0.05) preoperatively. GR number was not significantly different. Postoperatively, the Kd decreased. GR-ß mRNA expression was increased compared with controls (P < 0.05) and was decreased after surgery (P < 0.05). In patients with chronic hypocortisolism, GR- mRNA expression was increased, and receptor numbers were increased (P < 0.05), whereas GR affinity was normal. No changes were observed in patients undergoing a metyrapone test.
Conclusions: Cushing’s syndrome is accompanied by a reversible decrease in GR affinity, possibly related to an increased GR-ß expression, which may be a compensatory mechanism to GC excess. In chronic hypocortisolism, adaptive changes in GR status seem to occur at the level of GR number.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
B (1, 2).
Two GR isoforms were originally described: the functional GR- encoded by exons 2–8 and part of exon 9, which contains 777 amino acid residues, and the GR-ß where the 50 C-terminal amino acids encoded by part of exon 9 are replaced by 15 different amino acids encoded by part of exon 9ß. This difference renders GR-ß unable to bind GCs (3, 4). GR-ß has dominant-negative effects on hormone-induced GR- action (4, 5, 6), and increased expression of the GR-ß variant in steroid resistant states has been reported (7, 8, 9, 10). However, the physiological function of GR-ß is still the subject of debate (11, 12).
A third variant is the GR-P isoform, which has only 676 amino acids and is encoded by an mRNA encompassing part of intron 7, but lacking exons 8 and 9 (13, 14). The effects of the GR-P isoform in vivo are still unknown. However, we reported (15) that cotransfection of a GR-P expression plasmid with a GR- expression plasmid stimulated the GR--mediated effects of dexamethasone in a number of cell types (15), suggesting that GR-P might enhance GC bioactivity.
It is not known whether hyper- and hypocortisolism are associated with alterations in GR splice variant expression, which, in turn, could influence GR function. In vivo and in vitro studies suggest that cellular GR expression is down-regulated in response to increased GC exposure, providing negative regulation of GC action (16, 17). However, studies of GR number and affinity in peripheral blood mononuclear leukocytes (PBMLs) obtained from patients with endogenous hypercortisolism have shown varying results. Some studies found no changes in GR number (18, 19), whereas others showed (subtle) down-regulation of GR expression (20, 21). We previously reported decreased GR ligand affinity without changes in GR number in patients with pituitary-dependent Cushing’s syndrome (22), which was also found by others (19). No studies have been performed on changes in GR characteristics after reversal of GC excess by treatment of Cushing’s syndrome. Also, in patients with adrenal insufficiency, available data are conflicting, because both a normal and a decreased number of GC-binding sites have been reported as well as a normal and an increased GR affinity (18, 23, 24).
We investigated whether in vivo hypercortisolism (Cushing’s syndrome) leads to alterations in the relative abundance of the three splice variants in PBMLs, which, in turn, may modulate GC sensitivity. GR number and affinity as well as mRNA expression of GR splice variants were examined before and after treatment of patients with ACTH-(in)dependent Cushing’s syndrome. Furthermore, we examined GR characteristics in patients with chronic hypocortisolism due to hypopituitarism and the effects of acute (relative) hypocortisolism induced by metyrapone testing on GR isoform mRNA expression.
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Hypercortisolism.
Eighteen patients (age, 52 ± 12 yr; five men and 13 women) with the clinical and biochemical characteristics of Cushing’s syndrome were included. Of these patients 14 were found to have pituitary adenoma, three had ectopic ACTH-producing tumors (medullary thyroid carcinoma, paraganglioma, and pancreatic neuroendocrine tumor, respectively), and one had an adrenal adenoma. Patients were biochemically evaluated by assessment of cortisol diurnal rhythm (sampling at 0800, 1700, and 2200 h), 24-h urinary cortisol excretion (upper limit of normal, 850 nmol/24 h), and a 1-mg dexamethasone suppression test (cortisol cutoff value, 140 nmol/liter; (Table 1
).
|
Controls were 14 healthy volunteers (age, 32 ± 14; six men and eight women). All participants gave their informed consent, and this investigation was approved by the Erasmus Medical Center medical ethics committee.
Blood sampling
Patients with Cushing syndrome. In all patients with Cushing’s syndrome, baseline blood samples were collected in heparinized tubes before any treatment. All patients with pituitary adenoma underwent transsphenoidal adenomectomy with perioperative hydrocortisone infusion (100 mg/24 h). Postoperatively, hydrocortisone administration was tapered and was stopped on d 4. Postoperative evaluation of the pituitary-adrenal axis was performed by measurement of morning plasma cortisol and a metyrapone test on d 7. GR splice variants were measured in all patients at baseline and in six patients on d 6 after transsphenoidal adenomectomy. GR number and affinity were assessed in nine patients at baseline and in six patients on d 6 postoperatively. Of three patients with ectopic ACTH production, two were treated by laparoscopic bilateral adrenalectomy, and the patient with the cortisol-producing adrenal adenoma was treated by laparoscopic unilateral adrenalectomy. Blood was sampled at baseline and postoperatively after hydrocortisone administration was tapered to a physiological substitution dose of 20 mg/d for measurement of GR splice variants and GR number and affinity.
Patients undergoing metyrapone test and patients with chronic hypocortisolism. The metyrapone test was performed on the seventh and eighth postoperative days in adenomectomized patients who had not been treated with corticosteroids for 2 d before the test. The metyrapone test was performed as follows. From 0800 h on d 7, six dosages of 750 mg metyrapone were given at 4-h intervals. Blood samples were collected before the test and 24 h after the start of the test. At baseline and 24 h, plasma levels of ACTH and cortisol were measured. Plasma 11-deoxycortisol levels were measured at 24 h, with levels above 350 nmol/liter indicating normal pituitary-adrenal axis function (25). In all patients with biochemically established hypocortisolism, blood was sampled before hydrocortisone replacement therapy was started.
Isolation of PBMLs and RNA isolation
PBMLs were isolated as previously described (26). The PBML-enriched interphase contained less than 1% granulocytes. No systematic difference in cell yield was found between patients and controls. RNA was isolated using the High-Pure RNA isolation kit (Roche, Mannheim, Germany). RNA was quantified using the Ribo-Green RNA quantitation reagent (Brunschwig Chemie, Amsterdam, The Netherlands).
RT-PCR and real-time PCR assays
Eight hundred nanograms of total RNA was used in an RT reaction of 50 µl using 5 µmol/liter random hexamers and 200 nmol/liter oligo(deoxythymidine) primer in a first strand cDNA synthesis kit (TaqMan RT reagents, Applied Biosystems, Nieuwerkerk a/d IJssel, The Netherlands). Reactions lacking reverse transcriptase were run as controls for genomic DNA contamination.
For each of the GR splice variants, 2 µl of the resulting cDNA was amplified in separate real-time PCR assays on an ABI PRISM 7700 Sequence Detection System (Applied Biosystems). Assays were performed in a total volume of 25 µl containing 7.5 pmol of each primer and 5 pmol of the probe using the qPCR-core kit (Eurogentec, Liege, Belgium). After an initial denaturation at 95 C for 10 min, PCR was performed, consisting of 42 cycles consisting of 15-sec denaturation at 95 C and 1-min annealing and elongation at 60 C.
To detect the expression of the GR splice variants, we used a common upstream primer, 5'-TGTTTTGCTCCTGATCTGA-3', located in exon 6, as well as a common probe, 5'-FAM-TGACTCTACCCTGCATGTACGAC-tetramethylrhodamine-3', located in exon 7, for all isoforms. To discriminate among GR-, -ß, and -P, we used specific downstream primers. The sequences of these reverse primers were: rev-, 5'-TCGGGGAATTCAATACTCA-3', located in exon 9; rev-ß, 5'-TGAGCGCCAAGATTGT-3', located in exon 9ß; and rev-P, 5'-GTTTCTGCCATACCTATTTG-3', located in intron 7. These primer/probe combinations produced amplicons of 389 bp (GR-), 393 bp (GR-ß), and 365 bp (GR-P), respectively, at comparable efficiencies, as tested with plasmid DNA.
Apart from using the same amount (800 ng) of total RNA in all RT reactions, the results were further normalized by correcting them for the amount of hypoxanthine phosphoribosyltransferase mRNA, which is not affected by glucocorticoids in PBMLs. For this assay, the forward primer (12.5 pmol), 5'-CACTGGCAAAACAATGCAGACT-3'; the reverse primer (12.5 pmol), 5'-GTCTGGCTTATATCCAACACTTCGT-3'; and the probe (5 pmol), 5'-FAM-CAAGCTTGCGACCTTGACCATCTTTGGA-tetramethylrhodamine-3', were used. All primers and probes were obtained from BioSource Europe (Nivelles, Belgium).
Whole-cell dexamethasone binding assay
The number of GRs and their affinities for [3H]dexamethasone were measured in PBML in controls and Cushing’s syndrome patients before treatment and 6 d after surgery. The method used was described previously (26).
Hormone estimations
Plasma cortisol (reference value, 200–800 nmol/liter), urinary cortisol (determined in unextracted urine), and ACTH-(1–39) concentrations were measured by fluorescent immunoassays (Diagnostic Products Corp., Los Angeles, CA). Levels of 11-deoxycortisol were assessed by RIA as described previously (25), using an antiserum purchased from ICN Biomedicals (Costa Mesa, CA).
Statistical analysis
All results, relative amounts of GR isoform mRNAs, and biochemical determinations are presented as the mean ± SEM. Within the groups and comparing the groups with each other, statistical significance was calculated by ANOVA with a post hoc Bonferroni test. Correlations were tested using the Pearson correlation coefficient, and values before and after operation were analyzed using the paired samples t test. Differences were considered significant at P 
0.05.
For calculation of the relative amounts of each of the three GR mRNA variants, the comparative threshold cycle method was used (27), and the average value for GR- mRNA in controls was arbitrarily set at 100%.
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
protein in controls
Table 2 shows the levels of three GR mRNA splice variants and the dexamethasone binding capacity in controls. We found relative expression levels for GR-, GR-ß, and GR-P mRNAs of 100%, 0.08%, and 24%, respectively, with the level of GR- mRNA arbitrarily set at 100%. A significant correlation existed between the levels of all three splice variants (P < 0.05 in all cases; Fig. 1, A–C). Repeated measurements in a number of controls showed that both for receptor number and affinity and for mRNA expression levels, an intraindividual variation of approximately 20% existed (not shown).
|
|
Endocrine evaluation.
In all patients, hypercortisolism was biochemically established by the absence of a plasma cortisol diurnal rhythm, increased 24-h urinary free cortisol excretion, and insufficient suppression of plasma cortisol levels after oral administration of 1 mg dexamethasone (Table 1
). ACTH concentrations in these patients were high-normal or above normal (not shown). Of the patients with pituitary-dependent Cushing’s syndrome, six were postoperatively evaluated for GR characteristics on d 6 after hydrocortisone administration was stopped on d 4. Mean morning plasma cortisol levels decreased in all patients from 679 ± 67 nmol/liter at baseline to 310 ± 92 nmol/liter on d 6 postoperatively. According to these cortisol levels and the metyrapone test performed on d 7 (not shown), three patients were in remission (hypocortisolemic; cortisol on d 6, 106 ± 24 nmol/liter), and three patients had residual ACTH overproduction (cortisol on d 6, 513 ± 26 nmol/liter).
Expression of the GR splice variants and GR- protein in patients with Cushing’s syndrome.
The three GR splice variants were present in all patients with Cushing’s syndrome before and after operation (Table 2). There were no significant differences in the relative levels of the GR- and GR-P isoforms between controls and Cushing’s syndrome patients before operation. However, the relative level of GR-ß mRNA was significantly higher in Cushing’s patients before operation than in controls (P < 0.05; Table 2). All patients except one were surgically treated, and in nine cases [transsphenoidal adenomectomy, n = 6; (bi-)adrenalectomy, n = 3], complete pre- and postoperative data were available. Analysis of GR data before and after surgery showed that only the level of GR-ß mRNA had significantly decreased to normal values after operation (P < 0.05; Table 3). When this was evaluated separately in the group of patients with pituitary-dependent Cushing’s syndrome (n = 6), the difference remained, but was no longer significant. In the patients with nonpituitary-dependent Cushing’s syndrome, both GR- and GR-ß mRNA levels had significantly decreased after operation (P < 0.01; Table 3). There was no difference in the expression levels of the splice variants between the patients who were in remission and those who had residual ACTH overproduction. As in the controls, there was a significant correlation among levels of the GR-, GR-P, and GR-ß splice variants (Fig. 1). In all patient groups GR-P levels were considerably higher (2–64% of GR- levels in controls) than GR-ß levels (0.02–0.30%).
|
) in PBMLs of patients before surgery showed that the ligand affinity (1/Kd) was significantly lower in patients with Cushing’s syndrome compared with controls. GR number in patients was not different from that in controls. Furthermore, the mRNA levels of GR-P and GR-ß, but not those of GR-, were significantly correlated to the number of GR per cell (Fig. 2), but there was no correlation with the Kd. After surgery, the Kd decreased in six of nine patients (Table 3). With regard to the patients with pituitary adenoma, no significant differences were observed in postoperative Kd values in patients who were cured compared with patients with persistent hypercortisolism. The Kd decreased in all patients who underwent (bi-)adrenalectomy. Receptor number did not change significantly after pituitary or adrenal surgery (Table 3). There was no significant correlation between receptor parameters (mRNA levels, Kd, and receptor number) and the measured plasma and urinary cortisol values.
|
Expression of the GR splice variants in patients undergoing metyrapone test.
We measured the GR mRNA variants in seven patients who underwent metyrapone test after transsphenoidal adenomectomy for nonfunctional adenoma. In all patients, a normal reaction of the hypothalamo-pituitary-adrenal axis to metyrapone was found, as defined by an adequate increase in plasma 11-deoxycortisol levels (>350 nmol/liter). Administration of metyrapone resulted in an acute decrease in plasma cortisol levels with a concomitant increase in plasma ACTH concentrations (Fig. 3A). We did not find significant differences when comparing the mRNA expression levels of GR-, -P, and -ß before and after the test (Fig. 3B) or between this patient group and controls.
|
and GR-P mRNAs than in controls. However, only the increase in GR- mRNA was significant relative to controls and patients with nonpituitary-dependent Cushing’s syndrome (Table 2). At the level of the receptor protein, this was paralleled by an increase in the number of receptors per cell (P < 0.05 compared with controls; Table 2). Receptor affinity was not different from controls in patients with long-term hypocortisolism. |
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
(4); such effects were indeed shown in in vitro transfection experiments (5, 6, 10), and high expression of GR-ß in GC-resistant states has been reported (7, 8, 9, 10).
The GR-P variant was originally identified in a GC-resistant multiple myeloma cell line (13) and was shown to be the only GR variant expressed (mRNA and protein) in the DMS-79 cell line, which was also GC resistant (14). In contrast, transfection of a GR-P expression plasmid augmented the GC-induced action of cotransfected GR- (COS cells) or endogenous GR- (HeLa cells) (15). GR-P expression was found at high levels (5–20% of GR-) in a wide variety of hematological malignancies as well as in normal cells (15).
Assuming a concept in which GR-ß attenuates and GR-P facilitates GC action, it might be hypothesized that in a state of hypercortisolism, differential expression of GR isoforms, i.e. increased GR-ß and/or decreased GR-P expression, could counteract the effects of GC excess, whereas in a state of hypocortisolism, opposite changes could be anticipated. Therefore, we investigated whether the relative levels of GR-, GR-ß, and GR-P mRNA are subject to regulation by cortisol in vivo. Three patient groups were studied and compared with healthy controls: patients with Cushing’s syndrome, representing subjects subjected to long-term hypercortisolism; patients undergoing metyrapone testing, representing subjects exposed to acute (relative) hypocortisolism; and patients with (untreated) long-term hypocortisolism.
The relative expressions of GR- and GR-ß mRNA in PBMLs from healthy control subjects were similar to those reported previously (15, 29), with GR-ß mRNA present at 0.1% or less of GR- mRNA. PBMLs are a mixture of lymphocytes with 5–20% monocytes, and we could not quantify differences in expression of the GR variants between these cell types from our data. However, previous literature (29) showed relatively small differences between these types of leukocytes. Interestingly, GR-P mRNA was expressed at considerable levels (14–37% of the GR- mRNA) in PBMLs. Although no physiological function for this splice variant is known, it might play a role in the high GC sensitivity of lymphocytes.
In patients with Cushing’s syndrome, we measured GR- mRNA levels comparable to those in controls. In line with this, we did not observe changes in GR number in Cushing’s syndrome before or after treatment. This corresponds with previous studies (18, 23). In contrast, in vitro studies with various tissues and cell lines suggest a homologous GR down-regulation in response to GC (over-)exposure via inhibition of GR transcription (30), reduction of GR mRNA stability (16), or reduction of GR protein half-life (31). However, GR down-regulation does not seem to occur in chronic endogenous hypercortisolism, considering the normal GR number and GR- mRNA expression found in this study. Possibly, changes in GR expression do occur during the development of hypercortisolism, but disappear upon prolonged exposure. This may also explain the lack of association between cortisol levels and GR expression levels. It may be interesting to study GR expression from the start of treatment in patients treated with GCs. Previously, we have shown that in patients with pituitary-dependent Cushing’s syndrome, the GR shows a reduced affinity in the whole-cell dexamethasone binding assay (22). This was confirmed in the present study in patients with Cushing’s syndrome due to pituitary adenomas, ectopic ACTH production, or adrenal adenomas. Additionally, we found that GR affinity increased after hypercortisolism was (partly) reversed.
GR-ß mRNA levels were increased in patients with Cushing’s syndrome, were related to receptor number and were decreased significantly after surgery. The expression of GR-ß mRNA was approximately 1000 times lower than that of GR- mRNA. It was detectable in most samples, although the experimental variation was slightly higher than for the other variants, probably due to the higher threshold cycle for GR-ß mRNA. The GR-ß mRNA levels were lower than those reported by others (8, 15, 29, 32). This might be due to the use of different techniques: real-time-RT-PCR vs. Northern blot. The question arises of whether this low level of GR-ß expression can influence GR- function. There are reports that the protein half-life of GR-ß may be much longer than that of GR-, increasing cellular GR-ß availability (33). Indeed, at the protein level, the GR-/GR-ß ratio is generally lower than at the mRNA level in various tissues (34). In PBMLs a GR-/GR-ß protein ratio of 20:1 was measured (10). In contrast, others failed to demonstrate GR-ß protein in any tissue (29).
Several mechanisms can be proposed for the increased GR-ß mRNA expression in these patients. Cortisol itself might induce a dose-dependent increase in GR-ß expression, as observed in vitro (35, 36). Alternatively, cytokines such as TNF, IL-1, and IL-8, can enhance GR-ß expression in PBMLs and neutrophils (10, 37). In patients with Cushing’s syndrome, concentrations of IL-8 and IL-18, a member of the IL-1 family, are elevated (38) and may thus contribute to increased GR-ß expression.
Speculating that increased GR-ß expression could serve as a compensatory mechanism during hypercortisolism through its dominant-negative effects on GR- function, several possibilities exist. GR-ß may induce GC resistance by competition for coactivators or formation of heterodimers with GR (39), or effects on GR ligand affinity could be involved, considering the parallel changes in GR affinity and GR-ß mRNA expression in our patients. The exact mechanism underlying decreased GR affinity in Cushing’s syndrome remains to be elucidated and may also include changes in heat shock protein expression, which is known to influence GR ligand affinity (40).
We also measured the effects of metyrapone testing on GR isoform transcription. To exclude effects of preexistent hypocortisolism, only patients who appeared to have normal pituitary-adrenal axis function, i.e. with an adequate increase in 11-deoxycortisol levels, were studied. Short-term (24-h) hypocortisolism with concomitantly elevated ACTH levels did not induce significant changes in GR isoform mRNA expression. In contrast, in patients with chronic secondary hypocortisolism, GR- expression was significantly increased, with a concomitantly increased receptor number compared with controls and patients with Cushing’s syndrome. GR affinity as well as GR-ß and GR-P mRNA levels were within the normal range. This indicates that, in contrast to hypercortisolism, in long-term hypocortisolism, adaptive changes in GR status occur at the level of GR number rather than at the level of GR affinity. Possibly, the affinity of the GR can be down-regulated by posttranslational modification, but not up-regulated.
In conclusion, GR-, GR-ß, and GR-P mRNA were present in samples from controls, hypercortisolemic patients, and hypocortisolemic patients. In Cushing’s syndrome, hypercortisolemia led to a relative increase in GR-ß expression and a reduction in receptor affinity, whereas in hypocortisolemia, an increase in GR- mRNA expression was accompanied by an increase in the number of receptors per cell. GR-P mRNA levels did not differ significantly among the three groups. There was a highly significant correlation between the levels of the three mRNAs, both within groups and overall. This suggests that despite the changes observed in the relative expression levels of the three splice variants, neither hypercortisolism nor hypocortisolism has a major effect on the ratio at which these three splice variants are formed from their common precursor transcript. Apparently the primary transcript is processed at more or less fixed ratios into the three mature mRNAs, GR-, GR-ß, and GR-P.
|
Footnotes |
|---|
First Published Online August 23, 2005
Abbreviations: GC, Glucocorticoid, GR, GC receptor; PBML, peripheral blood mononuclear leukocyte.
Received May 11, 2005.
Accepted August 15, 2005.
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
- and ß-isoforms in human cells and tissues. Am J Physiol 283:C1324–C1331
and ß isoforms and type I 11ß-hydroxysteroid dehydrogenase expression in human skeletal muscle cells: a key role in the pathogenesis of insulin resistance? J Clin Endocrinol Metab 86:2296–2308This article has been cited by other articles:
 |
|
 |
|
  J. L. Chisa and D. T. Burke Mammalian mRNA Splice-Isoform Selection Is Tightly Controlled Genetics, March 1, 2007; 175(3): 1079 - 1087. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Russcher, V. A S H Dalm, F. H de Jong, A. O Brinkmann, L. J Hofland, S. W J Lamberts, and J. W Koper Associations between promoter usage and alternative splicing of the glucocorticoid receptor gene J. Mol. Endocrinol., January 1, 2007; 38(1): 91 - 98. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
